Fig 1.
Landscape composition of the Cecil Plains, Queensland, Australia, study site in the 2011/12 and 2013/14 cropping seasons.
Bt cotton is gray; mandated refuge is black; crops that provide breeding sites for both H. armigera and H. punctigera (source crops) are gray with dark-gray stripes while source crops for H. armigera alone are dark gray; non-breeding habitat (matrix) is light gray; native vegetation is stippled; and unknown habitat is white. In 2013/14 a few fields were planted with conventional cotton as mandated refuges within a more-extensive crop of Bt cotton (light gray with narrow stripes).
Table 1.
Baseline Helicoverpa preferences and seasonal availability for different habitat types.
Fig 2.
For model L, selection for resistance in H. armigera and H. punctigera under three scenarios about crop planting: (i) cotton (Bt and conventional) is only available in generations 2–4 (black lines); (ii) like (i) but 10% of cotton is also available for generation 5 (red lines and triangles); and (iii) like (i) but 10% of cotton is also available for generation 1 (green lines and circles).
The relative contributions of different crops and native vegetation for each of six years are based on field data from Cecil Plains. The top panels give K, the total available breeding habitat (Bt cotton, mandated refuges, and unmandated refuges) and the next-lower panels give Q, the proportion of breeding habitat that is refuge. Years are demarcated by vertical dashed lines. The 6-year patterns in Q and K are repeated in the simulations, so the values in first generation (first point on the left) arise from the values in the thirtieth generation (last point on the right). The next-lower panels are the log10 densities of eggs in refuges, measured before density-dependent mortality of larvae. The bottom panels give the strength of selection for resistance, Фt (Eq 1). For all simulations, F = 50, and habitat preferences are given in Table 1.
Fig 3.
Relative time to control failure for H. armigera (A,B) and H. punctigera (C,D) versus the proportion of cotton crop (Bt and conventional) that is either (A,C) available late to generation 5 or (B,D) available early to generation 1. Results are for model L (black lines), model A (dashed blue lines and +'s) and model I (turquoise dashed lines with x's). Time to control failure is given relative to the case of strict cotton cropping that is available to only generations 2–4, with time to control failure given when the allele frequency of resistance to Cry2Ab reaches 0.5. Habitat proportions are given by the landscape at Cecil Plains for 2009–2014, and the habitat preferences for each species are the same as Fig 2.
Fig 4.
For models L, A and I, the effects of variation in Q and/or K for four scenarios: (i) only Q varies, with Q1 = 0.05 and Q2 ranging from 0.05 to 0.2 (indicated on the horizontal axis labeled "Variation" which is scaled from 0 to 1 as Q2 goes from 0.05 to 0.2), while K1 = K2 = 1 (solid line); (ii) only K varies, with K1 = 1 and K2 in the ranging from 1 to 0.25 (indicated by values of "Variation" ranging from 0 to 1), while Q1 = Q2 = 0.05 (dashed line); (iii) Q and K vary asynchronously, with Q1 = 0.05 and Q2 ranging from 0.05 to 0.2, and K1 = 1 and K2 ranging from 1 to 0.25 (↑Q↓K, dotted line); and (iv) Q and K vary synchronously, as in (iii) but with high values of Q2 coinciding with high values of K2 (↑Q↑K, dot-dashed line).
(A-C) give the times to control failure (when resistance allele frequency reaches 0.5) for models L, A, and I, respectively. (D-F) give the strength of selection Фt (Eq 1) for four consecutive generations for models L, A and I, respectively, and for scenarios (i)—(iv). (G-H) give the densities of eggs (numbers/[QtKt]) in refuges before density-dependent mortality and averaged over both generations for models L and A, respectively. (I) The predicted relative time to control failure from the approximation giving Tc (Eq 2) and the observed time to control failure for all of the simulated runs for all three models shown in (A-C). Other parameter values are: F = 50, s1 = 0.01, s2 = 0.01, and h1 = h2 = 0.05.